Awards were announced May 2007

 

Michaela Biasutti and Adam Sobel, Columbia University
Mechanisms of 21st Century changes in Sahel Precipitation in the CMIP3 Climate Models

The CMIP3 models simulate contradictory pro jections for changes in rainfall over the African
Sahel over the 21st century. In some models, enhanced precipitation is predicted as a consequence
of enhanced land/sea temperature contrast and monsoon circulation (Haarsma et al.; 2005). In
others, the Sahel dries in response to the warming of tropical oceans, in analogy to the response
to a warm ENSO (Held and Lu; 2007). We propose to investigate these two mechanisms in the
pre-industrial, 20th century and A1B 21st century scenario integrations of all CMIP3 models.

The first mechanism requires that a hotter Sahara produce a stronger heat low, thus driving
a stronger monsoon circulation and heavier rainfall. Unfortunately, things need not be so simple.
For example, a hotter Sahara and a stronger circulation might at the same time bring dryer air
to overlay the monsoon layer over the Sahel and might cap convection, thus reducing rainfall. Or
the direction of influence between the Sahara low and Sahel rainfall might be opposite to what
hypothesized, with anomalies in Sahel rainfall causing a wave response that affects the strength
and the spatial extent of the Sahara low. We will use statistical analysis to diagnose, in each
model, the relationship between the strength of the Sahara low and other relevant quantities (for
example, surface temperature, energy fluxes at the surface, vertical structure of the boundary
layer, strength and vertical structure of subsidence, strength of convection in the Sahel and in
Asia, local circulation). Similarly, we will check how robust across models is the hypothesized
link between the land-sea temperature contrast and the strength of moisture convergence by the
monsoon circulation. More refined theories of the monsoon hold that surface temperature is not
the relevant quantity, but that one should look at the boundary layer moist static energy and
surface enthalpy fluxes. We plan to investigate these relationships as well.

The second mechanism for Sahel rainfall change holds that tropical SST warming leads to a
dryer Sahel. The hypothesis is that the main convective regions in the warming tropical oceans
set the tropospheric temperature of the entire tropics to a warmer moist adiabat profile. Regions
where moisture availability is limited might not reach the new threshold in boundary layer
moist static energy that would be necessary for deep convection to develop in this stabilized
environment, and might consequently become dryer. We will investigate whether long-term
drying can be achieved by the tropospheric stabilization mechanism, in particular whether and
to what extent a differential rise in moist static energy of land and ocean occurs on the long time
scales of global warming. If this is the case, then we can ask whether different model treatments
of vegetation and land surface processes can influence the time lag between land and ocean and,
ultimately, the Sahel response to a generalized warming.

This study will be diagnostic in nature, and will complement other theoretical and modeling
work on the mechanisms controlling mean tropical rainfall in GCMs that the investigators are
pursuing under different sources of support. We also anticipate that our ongoing collaboration
with Dr. Isaac Held (GFDL) will carry on in the proposed project, so that the results of our diagnostics
on several models and Held’s insights from more detailed analysis and experimentation
with a single GCM will feed back on each other.

Annalisa Bracco, Georgia Institute of Technology
ENSO and droughts over North America. The interdecadal variability of the SST forced signal.

Theoretical results (Trenberth and Guillemot, 1996), atmospherical general circulation
models (AGCMs) (Seager et al., 2005) and observational and proxy data (Cole and CooP, 1998)
show evidence that changes in the freSuency and amplitude of interannual variability in the
tropics can modulate occurrence and duration of droughts over North America.
Uarious mechanisms of atmospheric teleconnection between the tropical Pacific and North
America have been hypothesized to explain the tendency for droughts in presence of cool sea
surface temperature (SST) anomalies, i.e. for La-Nina-liPe conditions. @et the analysis of
instrumental records from the most recent decades emphasizes how incomplete our
understanding of those interactions really is.

With the goal of gaining a better understating of the role played by the Tropical Pacific during
the second half of the ZZ century and in future warmer climates in modulating occurrence and
duration of droughts in North America, the PI proposes to investigate the following Suestions:

• Which part of the circulation anomalies that create droughts over North America
  results from tropical SST forcing (ENSO \El Nino Southern Oscillation- in
  particular) and which part results from internal modes of the (chaotic) atmospheric
  dynamics?
• Would changes in the ENSO characteristics - variance and mean state in particular -
  affect the probability of drought occurrence and^or their duration and extension in the
  USA?

The first Suestion will be addressed by analyzing a large number of simulations collected as part
of the C20C international project. The potential predictability sPills of North American observed
rainfall will be assessed by systematically comparing different AGCMs outputs during the period
1950-1999 and separately during the 1950-1976 and 1977-1999 intervals. This will also allows
to assess the significance of changes in the atmospheric patterns associated with North America
precipitation before and after the 1976 climate shift in the Pacific. To investigate the second
point the PI will analyze a suite of AR4 IPCC coupled runs to test a) the role of ENSO variance,
direction of propagation of SST anomalies and persistency of SST and winds anomalies and b)
the role of the general warming in the tropics in modulating occurrence and duration of droughts
over North America.

Robert Burgman and Amy Clement, University of Miami - RSMAS
Past and Future North American Drought

Drought is a fact of life for American farmers in the central United States of America.
While droughts and pluvials are a part of the natural variability of North American
precipitation, periods of persistent drought like the “dustbowl” drought of the 1930’s can
be economically and environmentally disastrous. Though the persistent drought of the
1930’s was the single worst drought in the instrumental record, paleo-proxy data like tree
rings and lake sediments suggest there have been longer and even more severe droughts
in the central United States in the last 1000 years. The mechanisms behind these periods
of persistent drought remain unclear, however, recent research suggests that the tropical
Pacific Ocean may play an important role.

The purpose of the proposed work is to investigate whether this low frequency mode of
North American drought variability is simulated in the general circulation models that
simulate global climate. Analysis of fully coupled simulations of the 20th century will be
compared with observations and AMIP-style simulations where the atmospheric
component of the model is forced by observed SST’s. The analysis will then be applied to
simulations in which CO2 forcings are increased at 1% per year to simulate future
climate change. The result will be compared to the 20th century simulations in order to
understand how this mode of variability might change in the future.
Two paleoclimate simulations will also be included in the study. The first is an AMIP
style simulation in which SST’s are estimated for periods within the last 1100 years using
fossil coral records. This simulation is designed to test the ability of tropical Pacific to
simulate the persistent drought seen in the Paleo-proxy records. The Pacific Ocean
Global Atmosphere simulations with mixed layer ocean model - ML (POGA) use
prescribed SST’s in the tropical Pacific (30N – 30S) with a mixed layer ocean outside of
the Pacific. The second paleoclimate simulation uses a fully coupled climate system
model with varying solar, volcanic, and greenhouse gas forcings ( see figure 1). This
simulation is directly comparable to those coupled simulations for the past century and
offers an opportunity to analyze how the low frequency mode of North American drought
changes with longer timescale changes in the tropical Pacific.

The proposed work aims to advance our understanding of how a particular mode of North
American drought variability is simulated in the complex coupled models used to
simulate global climate. The work will improve our understanding of the ocean’s role in
North American drought and how persistent drought may change in the future. The
incorporation of extended AMIP-style and AOGCM paleoclimate simulations offers
insight into the impact of century scale variability associated with variations in solar and
volcanic forcings.

This study will be in collaboration with Richard Seager at the Lamont-Doherty Earth
Observatory of Columbia University and builds on results funded by NSF grant 66180W.

A. Capotondi, NOAA CIRES
Decade-Long Droughts in the Western U.S. and their Connection to Tropical Pacific Decadal Variability

Precipitation is of fundamental importance for life, and dry conditions persisting over several years in a given area can dramatically impact terrestrial ecosystems, agriculture, and the evolution of society. The western two thirds of the United States (here after referred to as the western U.S.) is an area subject to droughts that have persisted for several years to decades. These long periods of below-normal precipitation have been related to anomalously cold sea surface temperature (SST) conditions in the tropical Pacific, with a pattern similar to that of the Pacific Decadal Oscillation (PDO). We plan to use the climate simulations performed in support of the Intergovermental Panel for Climate Change (IPCC) Assessment Report 4 (AR4) and archived by the Program for Climate Model Diagnosis and Intercomparison (PCMDI) to examine whether state-ofthe-art climate models show indications of statistically significant correlations between western U.S. droughts and decadal SST anomalies in the tropical Pacific, and how that relationship may be modified by climate change.

The specific questions that we ask in this study are the following:
1. Are decade-long drought conditions found in the control and 20th Century
simulations of the AR4 models, and are they significantly related to tropical
Pacific decadal variability as a function of season?
2. What are the mechanisms responsible for tropical Pacific decadal SST
anomalies in the AR4 models? Are they in agreement with the leading
theories of tropical Pacific decadal variability?
3. Can we expect any change in tropical Pacific decadal variability, its
connection with Western U.S. precipitation, and drought duration and
frequency in different climate change scenarios?

We will answer these questions by examining long control simulations (whenever available), as well as 20th Century and climate change scenarios ensembles available from the PCMDI archive. Statistical analyses will be used to assess the relationship between western U.S. precipitation and tropical Pacific SST variability, while process-oriented approaches will be undertaken to identify the dynamical processes responsible for the decadal SST anomalies. The ability to understand the processes governing decadal variability in the tropical Pacific can help assess the level of predictability and the prediction lead time for decadal droughts over the western U.S.

Tsing-Chang Chen and Joseph Tribbia, Iowa State Univeristy
Droughts in the Central Plains of the United States

The potential causes for the geographic preference and remote forcing of warm-season droughts in the
U.S. Central Plains were identified:

1) The large-scale flow pattern of Central Plains droughts is characterized by a ridge in the western U.S. and a trough in the eastern U.S. This flow pattern resembles the spring-summer flow regime change across the U.S. continent which is accompanied by a drastic June-July decrease in the Central Plains rainfall maximum. This favorable continental-scale environment for droughts is established by the global spring-fall annual and the March-June semiannual modes which are much weaker than the winter-summer annual mode.
2) Without any special interannual forcing, the atmospheric circulation should only undergo a regular seasonal variation in response to the seasonal cycle of solar heating. Thus, to identify possible interannual forcing and the atmospheric response to this forcing, the regular seasonal cycle should be removed. Applying this approach, a cross-Pacific short-wave train emanating from an interannual rainfall anomaly center over the western subtropical Pacific appears to couple with the Central Plains drought flow pattern across the U.S. continent.

Using these findings, a study is proposed to pursue two research tasks: 1) to evaluate whether the seasonalcycle modes forming favorable environments of Central Plains droughts are well simulated by global (SMIP2 and DEMETER) and regional (NARCCAP) climate models, and 2) to test whether simulated Central Plains droughts are stimulated by the western subtropical forcing. The goal of the proposed study is to introduce new mechanisms leading to Central Plains droughts.

Kerry Cook and Edward Vizy, Cornell University
Hydrodynamics of the Caribbean Low-Level Jet and its Relationship to Drought

Output from coupled atmosphere/ocean global climate models and NCEP’s North American regional reanalysis will be analyzed to improve our understanding of the hydrodynamics of the Caribbean low level jet and its relationship to summer drought over North and Central America, and to evaluate our ability to capture this relationship in models. Past studies suggest that drought over the central United States is associated with a decrease in the poleward transport of moisture from the Gulf of Mexico by the Great Plains lowlevel jet. The meridionally-oriented Great Plains jet is located just north of the zonally-oriented
Caribbean low-level jet, and they often appear to be connected in monthly mean climatologies. However, the basic hydrodynamics of the Caribbean jet and how it varies during drought are not well known.

We will evaluate the role of the Caribbean low-level jet in supplying moisture to the central United States through interactions with the Great Plains jet, and in supplying moisture directly to Central America, contrasting the climatology with times of drought. Atmospheric and surface moisture budget analyses, as well as thermodynamic balances, will be diagnosed in the reanalysis and in the GCM output from 20th Century IPCC AR4 integrations in five models. This will allow us to better understand the hydrodynamics of drought and also to assess our ability to capture the proper mechanisms of drought in the models.

Eric DeWeaver and David Lorenz, University of Wisconsin
The Role of Land-Atmosphere Coupling in Perpetuating Drought

The objective of the proposed research is to understand and determine the role of land-atmosphere coupling in perpetuating drought in both climate models and in observations. Because soil moisture anomalies have large persistence, regions with strong(positive) soil moisture/precipitation coupling will have this soil moisture persistence imparted on the precipitation variability and will act to perpetuate both droughts and wet spells. Despite its importance in perpetuating drought, the effect of soil moisture on precipitation is extremely variable among current climate models, indicating a pressing need to understand the role of land-atmosphere coupling in drought.

For this study, we will analyze the climate models participating in the Intergovernmental Panel on Climate Change (IPCC) fourth assessment report (including the CCSM, GFDL and GISS climate models) and the high-resolution model datasets that are part of the North American Regional Climate Assessment Program (NARCAP). For observations we will use the North American Regional Reanalysis (NARR), a retrospective Variable Infiltration Capacity (VIC) model simulation of the land surface hydrological cycle, global data from the Global Land Data Assimilation Systems (GLDAS) project and data from the Southern Great Plains Atmospheric Radiation Measurement (ARM) site. We will use lagged correlation/regression analysis to quantify the effect of local soil moisture and evaporation on later precipitation for points over the entire globe. Preliminary analysis suggests that the lagged correlation analysis yields results similar to the sophisticated ensemble technique of Koster et al. (2002). This is good news because an important advantage of the lagged correlation analysis is that one can compare land/atmosphere feedbacks in climate models to observations.

We will define three separate drought indices based on precipitation, soil moisture or runoff. These three variables are used in the literature as the basis for defining "meteorological", "agricultural" and "hydrological" droughts, respectively. For each month, location and dataset, we will calculate monthly percentiles of occurrence for the
entire period of record. We will use the Climate Prediction Center (CPC) drought scheme, which classifies droughts based on percentiles. Finally, we will relate the strength of land-atmosphere coupling to the duration and intensity of droughts. Important questions this research will attempt to answer are: Are climate models with strong land/atmosphere coupling more likely to have longer and more intense droughts than models with weak coupling? Are regions of strong landatmosphere coupling more likely to suffer droughts of long duration compared to other
regions? Are climate models with large land-atmosphere coupling more likely to experience longer and more intense droughts in response to climate change?

Ian Ferguson, John Dracup and Phil Duffy, University of California, Berkeley
Stochastic Characteristics of Drought and Surface-Atmosphere Drought Forcing in Coupled GCMs

Droughts are the nation’s costliest natural disasters, with average annual impacts of $6-8 billion in the
United States alone (FEMA 1996); however, the physical and dynamical methods that lead to drought and the potential impacts of climate change on drought characteristics are poorly understood. This study will address two primary research questions:
• To what degree do surface-atmosphere interactions drive the stochastic characteristics of
drought, including drought duration, magnitude, and recurrence, over selected regions?
• How will anthropogenic climate change impact the physical and dynamical mechanisms that
lead to drought, and subsequently the stochastic characteristics drought?

Stochastic characteristics of drought will be evaluated globally based on observed climate data and all 19 transient 20th century coupled climate simulations in the WCRP CMIP3 data archive. A combined composite and multiple regression analysis will be used to objectively identify statistically significant patterns of antecedent and contemporaneous sea surface temperature (SST) and soil moisture (simulations only) associated with drought events. The role of surface-atmosphere forcing in the onset, persistence, and recession of drought events will then be assessed based on the amplitude and robustness of surface anomaly composites and multiple regression coefficients. Results will be compared between observations and simulations, and between models. Finally, the analysis will be repeated for all paired pre-industrial control and forced stabilization simulations in the CMIP3 data archive; potential impacts of climate change will be evaluated by comparing the drought characteristic and surface-atmosphere drought forcing between the control and forced stabilization simulations for all models.

The application of climate models to drought analysis is relatively recent, and few studies have investigated drought characteristics and forcing mechanisms in fully coupled GCMs. By improving our understanding of the stochastic nature of drought and the physical mechanisms that lead to drought events, the proposed analysis will help to improve drought planning and management, and ultimately reduce drought-related impacts to sensitive human and environmental systems.

Lisa Goddard, International Research Institute for Climate
Diagnosing El Nino induced tropical droughts in seasonal forecasts and climate change projections

El Nino brings widespread drought to the tropics, including Mexico and northern South America.
Stronger or more frequent El Niño events in the future will exacerbate drought risk in these highly vulnerable areas. Even if the frequency and intensity of El Niño events do not increase in the 21st century, more generalized warming of the tropical Pacific may still produce a tropical teleconnection resembling that
associated with present-day El Niño conditions.

We propose to evaluate the patterns, spatial extent, and severity of El Niño induced tropical droughts
during a control period in the 20th century in seasonal forecasts, which have updated realistic initial
conditions but fixed greenhouse gases (GHGs), and climate change projections, which have realistic GHG
evolution but no observational updates. We will then examine the projected changes in the strength of the
identified patterns of tropical drought in the 21st century runs. This research will address the following
questions:
1) How well do coupled models simulate the pattern and intensity of tropical droughts associated with
El Niño during the 20th century?
2) What are the primary differences of El Niño variability and change between seasonal prediction
models and climate change models in the 20th century?
3) To what extent do patterns of interannual precipitation variability project on 21st century precipitation
trends, and what is the spatial signature of the remaining trends?

This work will consider the strengths and weaknesses of seasonal predictions and climate change
drought projections, in particular how the models respond to large scale changes in sea surface temperature
(SST) such as those due to ENSO or to climate change. This is important, because if the impact on
precipitation fields from anomalous SST forcing on seasonal timescales is poorly simulated in climate
change projections, then the impact on such precipitation fields from increasing greenhouse gas forcing
cannot be trusted either. More optimistically, if the models contain robust information, and differences are
due mainly to systematic biases, then it may be possible to spatially recalibrate the predictions/projections
and provide more confident estimates of near-term and longer-term drought risk within the tropics.

Kristopher Karnauskas and Antonio Busalacchi, ESSIC - University of Maryland
Understanding Tropical-Subtropical Forcing and Predictability of Long-term North American Drought in Coupled Models

U.S. CLIVAR has initiated a Drought in Coupled Models Project (DRICOMP) to
encourage diagnostic research into the physical mechanisms of drought and evaluate its
simulation in existing coupled models. The common thread among a growing body of literature
is that ENSO is a fundamental driver of global drought variability, and La Niña-like conditions
play a key role in the circulation anomalies leading to North American drought. An important
yet poorly understood process with strong implications for understanding and predicting longterm
North American drought is how the coupled ocean-atmosphere system maintains persistent
cool conditions in the equatorial Pacific Ocean.The questions to be addressed in this proposal are aligned with the objectives of DRICOMP: (1) how are persistent cool episodes in the equatorial Pacific Ocean represented in today’s state-of-the-art coupled climate models, (2) through what physical mechanisms do such conditions lead to persistent North American drought, (3) how can an understanding of such
questions serve to extend the predictive lead time for long-term North American drought, and
(4), how do the mechanisms and prospects for predictability depend on the essence of a climate
in transition?

The work described in this proposal takes a two-pronged approach. First, a selection of
coupled models as determined from the details of their ocean component will be used to
characterize and explain the representation of persistent cool equatorial Pacific conditions.
Secondly, long-term North American drought variability (in terms of PDSI) and its relationships
with the equatorial Pacific, including how that appears to depend on anthropogenic forcing, will
be analyzed in the full suite of CMIP3 climate models. The specific contribution of the proposed
work to the problem of predictability will be an extended lead time, since we will ascertain (1)
how the remote forcing field itself (i.e. low-frequency evolution of equatorial Pacific SST) is
modulated, and (2) the coherence between that signal and the predictand (PDSI).

Brad Lyon, International Research Institute for Climate
Investigating the joint occurrence of summer drought and heat waves in climate change projections

 

Michael McPhaden and Dongxiao Zhang, NOAA PMEL
Ocean's Role on Tropical SST and Mid-Latitude Droughts in Climate Models

 

Xiao-Wei Quan, Martin Hoerling and Jon Eischeid, NOAA CIRES
Quantifying Drought in CMIP Simulations

This proposal seeks to perform a quantitative diagnosis and evaluation of drought behavior in the CMIP simulations. The focus will be on the statistical distribution of drought metrics including their severity, frequency of occurrence, duration, and spatial extent in the CMIP simulations of the 20th century with direct comparison to observations. Various drought indices including the conventional Palmer Drought
Severity Index (PDSI), the self-calibrating PDSI (SC-PDSI), and the Standardized Precipitation Index (SPI) will be used to ensure the robustness of our evaluation to the choices of drought definition.

Alfredo Ruiz-Barradas and Sumant Nigam, University of Maryland
SST – North American Drought Links in 20th Century Simulations and 21st Century Climate Projections

The proposal seeks to advance understanding of the genesis of long-term drought over North America, especially, continental US, from analysis of its concurrent and antecedent SST links. The observed SST links will serve as target in core assessments of drought mechanism (especially, initiation) in IPCC and CMIP3 simulations of 20th century climate. The assessment is critical as only mechanisms, and not events (e.g., the 1930 Dust Bowl drought), can be validated in coupled climate simulations. The veracity of SST–drought links in 20th century simulations can generate confidence in model-based projections of US drought incidence and intensity in a warming scenario.

To facilitate examination of a large number of model simulations and projections, a seasonal precipitation-based drought index is devised to capture low-frequency hydroclimate variability. The devised index is well correlated with the Palmer Drought Severity Index, which is widely used in US drought analysis and monitoring; the correlation is !0.8 over the Great Plains. The salient features of the proposed analysis are
• Development of precipitation-based drought index, as precipitation is universally archived
• All-season analysis of drought-SST links, given the soil-moisture sequestration in winter
• Evolution-centric depiction of combined SST–drought variability, from extended-EOF analysis
• Identification of recurrent modes of SST–drought variability, leading to quantification of basin
influences and insights into extreme event occurrence
• North American droughts in a warmer climate: New modes of variability or reordered preferences of existing ones in model projections?

We are well positioned to undertake this drought analysis given our ongoing focus on Great Plains hydroclimate variability in observational and model simulation data sets, including NCAR, NASA and GFDL ones. We are currently engaged in reconstruction of US droughts from Pacific SST anomalies. The PI and the co-I have also each lead CLIVAR-CMEP projects, to successful completion, and publications. The PI’s project, titled Diagnosis of North American Hydroclimate Variability in Coupled Model Simulations, was on a related, but more interannually focused, theme.

Sang-Ik Shin, NOAA CIRES
Changes in the impacts of tropical SSTs on the global and regional hydroclimate

Ning Zeng, University of Maryland
Seasonal Cycle of Drought in Coupled Climate Models and its Implication for the Hydro-ecosystem

The severity of drought can be better understood if its seasonal cycle is revealed. For example, drought
impact is much larger when it occurs during the dry season than wet season. We propose to analyze the seasonal cycles of precipitation change in the CMIP3 models, in a way significantly beyond the typical analysis of total precipitation change. The 20th century control climate will be analyzed and compared with observations to assess the realism of seasonal cycle as well as the annual mean simulated by these models. In doing so, a region-dependent criterion will be developed to serve as a basis for evaluation of predicted climate change. Then, precipitation changes from the 20th century to the future from all the archived models will be analyzed. Their seasonal cycle, especially the dry season behavior will be emphasized, and regions with the more robust signals will be classified by seasons. In doing so, we will provide insight into the likelihood and mechanisms of future droughts. While the analysis is inherently global, we will focus on three regions: the Mississippi basin, the Sahel, and the Amazon which represent three major climatic regimes and have great economic and environmental importance.

In addition, the projected future changes in precipitation, temperature, soil moisture and other relevant
variables from representative CMIP3 models will be used to drive a coupled land-surface and dynamic
vegetation model VEGAS. The ecosystem and water cycle sensitivity to different characteristic changes in
seasonal cycles will thus be assessed. The simulated soil moisture and vegetation state (such as leaf area index) will be compared to traditional drought indices such as precipitation anomaly, the Palmer Drought Severity Index (PDSI), and the Standardized Precipitation Index (SPI). This will shed light on the important question of how adequate the traditional drought indices are for the purpose of characterizing/predicting future changes of the hydro-ecosystem. This will also assess the potential ecosystem risks due to future drought.